Questions worth building around.
USCO research explores robotics, AI, embedded intelligence and practical technology through experiments and prototypes. Each line of enquiry is stated as a question with a method, and labelled with how far it has got.
Open questions.
Can biosignals drive a limb reliably?
Acquiring EMG, extracting a stable control signal and mapping it to a grip. Status: internal R&D with a functional demonstrator.
How much intelligence fits on a low-power device?
Small vision and language models on edge hardware, measured for accuracy, latency and power. Status: ongoing, with field-trial evidence for vision.
What keeps remote sensing alive when power and links fail?
Solar budgets, mesh topologies and store-and-forward. Status: field-tested in the farm system.
Can the home stay smart when the internet goes down?
Local AI and local control for the smart home. Status: in development.
How do world-model ideas help physical machines plan?
Exploring how learned models of the world can support planning in robots. Status: exploratory reading and small experiments.
How we run an experiment.
Hypothesis first
Written before the test, with a success threshold.
Smallest fair rig
Just enough hardware to answer the question.
Repeat and log
Failures recorded alongside successes.
Report plainly
What was tested, what was found, what it does not show.